PANDA Meeting @ GSI 13. December 2006

PID TAG

Georg Schepers

PANDA

Technical Assessment Group PID

Status Report

G. Schepers for the PID TAG

GSI

PID TAGPID TAG

http://wiki.gsi.de/cgi-bin/viewauth/Pandatagpid

PANDA Meeting @ GSI 13. December 2006

PID TAG

Georg Schepers

PID @ PANDA

• PANDA @ FAIR

• Global Requirements for PID

• PID TAG work

- Kinematic Studies

- Performance – Separation Power - Efficiency

- Global PID analysis

• Conclusion and Outlook

GSI

PID TAGPID TAG

PANDA Meeting @ GSI 13. December 2006

TAG-PID

Georg Schepers

pp- and pA-annihilation

GSI

PANDA @ FAIR

• wide range of momentum - p momentum: 1.5-15 GeV/c - mass range: 0-6 GeV/c • highest intensity - event rate 2 x 10 /s - up to 10 reaction products

• rich physics program

7

2

PANDA Meeting @ GSI 13. December 2006

TAG-PID

Georg Schepers

PANDA (PID) Detectors

GSI

PID TAGPID TAG

PANDA Meeting @ GSI 13. December 2006

PID TAG

Georg Schepers

• Subject

- Requirements from physics

- Evaluate potential of each subsystem

- Matching of systems

• Deliverables

- Definition of global PID scheme

- Optimized set of detectors and parameters

• Time Frame

- One year from asap ( PANDA Meeting @ DD 9.March 2006)

GSI

Global PID Requirements

PANDA Meeting @ GSI 13. December 2006

Georg Schepers

PANDA Technical Assessment Group PID actual

C. Schwarz, G. Schepers Chair

B. Kopf / R. Nowotny Barrel Calorimeter

R. Kaiser / B. Seitz Forward Cherenkov

O. Denisov / M. P. Bussa / M. Maggiora Muon counter

K. Föhl / P. Vlasov / M.Ehrenfried/ B.Seitz Cherenkov Counter

J. Smyrski / P. Hawranek / O. Wronska Forward Calorimeter

Q. Weitzel / B. Ketzer / S. Neubert TPC

(Carsten Schwarz /Aida Galoyan ) ToF

K. Peters Detectors

GSI

PID TAGPID TAG

PANDA Meeting @ GSI 13. December 2006

TAG-PID

Georg SchepersGSI

PID TAG WorkPID TAG Work

PANDA Meeting @ GSI 13. December 2006

TAG-PID

Georg SchepersGSI

Kinematics of the PANDA decay channelsKinematics of the PANDA decay channels

• Important input for common plots for Separation Power

• Solving open questions: - which quality for separation power do we need for different spatial regions ? - do we need a Forward Cherenkov ?

• MC truth studies for the PID - distinction between the different particles IDs

- distinction between different momentum regions : < 1 GeV/c : PID via dE/dx and/or ToF (and TPC) > 1 GeV/c : PID via Cherenkov detectors

- distinction between different spatial regions θ : 0° - 5° : FS- θ : 0° - 5° : Forward Spectrometer (FS) θ : 5° - 20° : forward part of the Target Spectrometer (TS) θ : 20° - 90° : forward part of the barrel region in the TS θ : 90° - 170° : backward part of the barrel region in the TS

PANDA Meeting @ GSI 13. December 2006

TAG-PID

Georg SchepersGSI

Kinematics of the PANDA decay channelsKinematics of the PANDA decay channels

DPM @ 2.0 GeV/cDPM @ 15.0 GeV/c

PANDA Meeting @ GSI 13. December 2006

TAG-PID

Georg Schepers

Do we need a Forward Cherenkov?

GSI

PID TAGPID TAG

PANDA Meeting @ GSI 13. December 2006

TAG-PID

Georg Schepers

Forward Cherenkov

GSI

PID TAGPID TAG

Forward RICH angles: 0-5/10deg

pions: 0.5 - 16 GeV/ckaons: 2 - 16 GeV/c

protons: 2 - 20 GeV/cthickness 1.2m =~ 0.11 Xo

PANDA Meeting @ GSI 13. December 2006

PID TAG

Georg Schepers

6 Parameter Analysis Endcap DIRCEndcap DIRC

GSI

Performance – Separation Power - Efficiency

K – π - Separation Power:σ - lines as function of θ and p1

2

34

PANDA Meeting @ GSI 13. December 2006

TAG-PID

Georg Schepers

Performance – Separation Power - Efficiency

Example : Endcap DIRC

GSI

Distribution and Resolution of β

Separation Power ( θ, p)

Efficiency K

π

PANDA Meeting @ GSI 13. December 2006

PID TAG

Georg SchepersGSI

Barrel DIRC

• Pin hole

• Focussing DIRC

• Time Of Propagation 3D DIRC

PANDA Meeting @ GSI 13. December 2006

PID TAG

Georg SchepersGSI

DIRC ParametrizationDIRC Parametrization

Separation Power =l m1²-m2² l / (2 p² σ-track(θ) sqrt(n*n-1)) Path length [cm] Photons produced

Separation Power vs p vs θ

p θθ

θ

p

pEfficiency

p θ

θ

p

p

PANDA Meeting @ GSI 13. December 2006

TAG-PID

Georg SchepersGSI

TPCTPC

PANDA Meeting @ GSI 13. December 2006

TAG-PID

Georg Schepers

TPC Parametrization dE/dx

GSI

TAG-PIDTAG-PID

Seperation Power = 2 l I1-I2 l / (sigmaI1/I1 + sigmaI2/I2)

I is dE/dx of the resp. particle. (resolution 5%)

PANDA Meeting @ GSI 13. December 2006

TAG-PID

Georg Schepers

GSI

(Barrel) EMC electron ID Studies(Barrel) EMC electron ID Studies

• Full simulation chain in BaBar like software - G4 simulation with the complete detector, digitization, full reconstruction for the EMC - single particles between 0.05 … 6.0 GeV/c and cos() =-0.7 … 0.7

- 10k e+

- 10k +

• Electron can be identified via- E/p (E: energy deposit of the cluster; p: reconstructed momentum of the track)- shower shape of the cluster

• Studies based on- complete EMC reconstruction

- reconstructed energy deposit of the cluster- reconstructed shower shape of the cluster (Zernike moments)- events with only one cluster - tracking not taken into account yet

PANDA Meeting @ GSI 13. December 2006

TAG-PID

Georg Schepers

GSI

(Barrel) EMC electron ID Studies – Training Neuronal (Barrel) EMC electron ID Studies – Training Neuronal NetworkNetwork

e π Κ+ + +

eπΚ

+++

PANDA Meeting @ GSI 13. December 2006

TAG-PID

Georg Schepers

GSI

(Barrel) EMC electron ID Studies – Results(Barrel) EMC electron ID Studies – Results

e+

+

e+

+

all all all

cut cut cut

Κ+

PANDA Meeting @ GSI 13. December 2006

TAG-PID

Georg Schepers

GSI

(Barrel) EMC electron ID Studies – Results(Barrel) EMC electron ID Studies – Results

PANDA Meeting @ GSI 13. December 2006

TAG-PID

Georg Schepers

GSI

(Barrel) EMC electron ID Studies - Outlook(Barrel) EMC electron ID Studies - Outlook

• Electron ID via e/p and shower shape seems to work

• PID for p < 0.8 GeV/c to be improved

• need of precise test beam data and Monte Carlo models

• optimal training of the neuronal network necessary

• combination with other detectors (e.g. dE/dx, tof, cherenkov)

• studies with different crystal sizes

PANDA Meeting @ GSI 13. December 2006

PID TAG

Georg SchepersGSI

Forward Calorimeter

Up 5 GeV/c down 1 GeV/c

The Forward EMC can be treated in the same way as the Barrel EMC

The groups from Bochum and Crocow work already on common simulations

PANDA Meeting @ GSI 13. December 2006

PID TAG

Georg SchepersGSI

Forward Calorimeter

Objectives of the Shashlyk Calorimeter π/e discrimination energy, position and time measurement

TPR Layout of a Shashlyk module

100 ps / sqrt(E) Time resolution

4.% / sqrt(E) Energy resolution

 

110 x 110 mm2Lateral dimensions

20 X0 Radiation depth

59.8 mm Effective Moliere radius

0.275 mm Lead layer thickness

1.5 mm Scintillator layer thickness

378Number of layers

PANDA Meeting @ GSI 13. December 2006

PID-TAG

Georg SchepersGSI

Barrel ToF

PANDA Meeting @ GSI 13. December 2006

PID-TAG

Georg Schepers

Muon Counters

GSI

PID-TAGPID-TAG

Different Strategies for muon identification in different momentum regions must be envisaged for muon detectors only out side the iron yoke:

E mu < 1GeV : muons can not reach the muon trackers => identification Barrel DIRC, Barrel EMC E mu > 1 GeV : identification feasible by means of muon trackers (gaseous detectors)

Basic detector features The bacis detector features for the muon trackers will be:

Ar, CO2, CF4 Gas mixture

25 nsec Double hit resolution

100 nsec Maximal drift time

1 cm Single plane space resolution (pitch size)

2-4 nsecTime resolution

Layer # 27

Layer # 7

PANDA Meeting @ GSI 13. December 2006

PID-TAG

Georg Schepers

Muon Counters

GSI

PID-TAGPID-TAG

PANDA Meeting @ GSI 13. December 2006

PID-TAG

Georg Schepers

Global PID Analysis

GSI

PID-TAGPID-TAG

• Evaluation for Global PID analysis for PANDA already started - number of PID variables ? - number of PID stages ? - combinations of detectors…

• Approches - Maximum Likelihood - method from other experiments compared (HERMES, Belle)

• First Conclusions taken: - Need of probability functions from test beam and simulations from all PID detectors and some traking input (e.g. p, θ) - Likelihood approach seems to be flexible enough - Analysis with Neuronal network interesting either as Global PID Analysis or as input for Likelihood method

PANDA Meeting @ GSI 13. December 2006

PID-TAG

Georg SchepersGSI

Conclusion for the PID TAG

• Separation Power parametrization and Efficiency determination prepared for some detectors for the others in progress

• Kinematic Plots for Benchmak channels available - background studies has to be done

• Questions : - Forward Cherenkov - positiv - TPC helpful as PID detector especially in low mommenta regions

• Concepts for Global PID Analysis under evaluation

PANDA Meeting @ GSI 13. December 2006

PID-TAG

Georg Schepers

GSI

Outlook

Next Steps:

• Need of precise detector tests and elaborated MC methods

• Need the Matching of - detector response functions - PID detectors with output from the tracking detectors

PANDA Meeting @ GSI 13. December 2006

TAG-PID

Georg Schepers

C. Schwarz, G. Schepers Chair

B. Kopf / R. Nowotny Barrel Calorimeter

R. Kaiser / B. Seitz Forward Cherenkov

O. Denisov / M. P. Bussa / M. Maggiora Muon counter

K. Föhl / P. Vlasov / M. Ehrenfried Cherenkov Counter

J. Smyrski / P. Hawranek / O. Wronska Forward Calorimeter

Q. Weitzel / B. Ketzer / S. Neubert TPC

(Carsten Schwarz/Aida Galoyan ) ToF

K. Peters Detectors

Thanks for important input:

M. Düren, K. Götzen, Th. Hennino a.o.

GSI

PANDA Technical Assessment Group PID